Detection and Prevention of False Data Injection Attacks in the Measurement Infrastructure of Smart Grids

Abstract

The smart grid has become a cyber-physical system and the more cyber it becomes, the more prone it is to cyber-attacks. One of the most important cyber-attacks in smart grids is false data injection (FDI) into its measurement infrastructure. This attack could manipulate the control center in a way to execute wrong control actions on various generating units, causing system instabilities that could ultimately lead to power system blackouts. In this study, a novel false data detection and prevention paradigm was proposed for the measurement infrastructure in smart grids. Two techniques were devised to manage cyber-attacks, namely, the fixed dummy value model and the variable dummy value model. Limitations of the fixed dummy value model were identified and addressed in the variable dummy value model. Both methods were tested on an IEEE 14 bus system and it was shown through the results that an FDI attack that easily bypassed the bad data filter of the state estimator was successfully identified by the fixed dummy model. Second, attacks that were overlooked by the fixed dummy model were identified by the variable dummy method. In this way, the power system was protected from FDI attacks.